| Abstract | [Introduction]
It has been suggested that a current limitation in our understanding of the mechanisms of motor control is the limited knowledge regarding the structure of the central nervous system as a multivariable controller (Ito, 1982). In attempting to understand the control of speech movements and the aberrations associated with speech movement disorders, the problem is compounded because the multiple movements of this motor act are subordinate to an overall communication goal. This is illustrated by the presence of motor equivalence in which speech motor actions are implemented in a complex, but systematically variable manner (cf. Abss, 1986; Gracco & Abbs, in press, b; Hughes & Abbs, 1976; MacNeilage, 1970). In response to the apparent variability in the movements and muscle activities associated with speech production, experimental effort has been directed toward the identification of invariant relations in the speech production process. The search for invariance is based on the assumption that stereotypic or invariant motor actions reduce the control complexity of a motor task. While the search for movement invariance has provided some general insight into motor control considerations, many issues remain unresolved (cf. Perkell, 1986; Stein, 1982 for reviews).
A major factor often overlooked in the search for and identification of invariant or simplifying relations is the mechanisms by which such relations are mediated. For example, one hypothesized invariant relation for limb and speech movements involves the scaling of a base velocity profile which accounts for the kinematic changes associated with movements of different amplitudes and durations (Atkeson & Hollerbach, 1985); Munhall, Ostry & Prush, 1985; Ostry & Cooke, in press). One consequences of focused attempts to uncover underlying control principles reflected in invariant relations has been in the limited consideration of sensory information in the motor control process. It is often overlooked that scalar adjustments could not be accomplished without sensory information regarding the relative position of the structures prior to movement as well as an estimation of the distance to be moved.
In this chapter, it will be suggested that sensory information is used in multiple ways during the planning, programming, and execution speech movements. Without considering the influence of sensory information and the underlying dynamic process, invariance may be relegated to an interesting observation. Variance, on the other hand, provides information on the underlying task-specific goals a well as insight into the mechanism underlying hypothesized invariant relations. Perhaps more can be gained in our understanding of the speech motor process if we attempt to account for the sources of variability as well as continue the search for invariant articulatory relations. In the next section, a perspective will be presented which briefly considers the importance of sensory information in the control of speech movements. More detailed considerations have been presented elsewhere (Borden, 1979; Gracco, 1984; Gracco & Abbs, in press). |